[Fwd: LFS Mirror]

Anderson Lizardo andersonlizardo at yahoo.com.br
Sun Nov 9 09:19:41 PST 2003

On Sat, 08 Nov 2003 22:35:50 -0700, Gerard Beekmans wrote:
> Eventually I'd like to be able to comment out mirrors quickly. If we
> have a flat mirrorfile (just one mirror URL per line without the
> /lfs/news.html tagged on) it'd be a simple matter to comment out a
> mirror line by, for example, putting a capital X as the first character
> of that line (Xhttp://stalemirror). The script that builds the
> index.html files can parse this file line by line and filter out lines
> that match the "^X.$" regular expression.

A plain text file with only the mirror URL is not possible AFAIK,
because we need also to maintain the mirror location and the link

So I suggest we use a CSV (Comma-Separated Value) database. It's a
simple plain text file, with one entry per line. Eg:

# Mirror URL, Location, Link Bandwidth
http://lfs.oregonstate.edu,"Corvallis, Oregon, USA",310 Mbps
http://www.linuxfromscratch.org,"Calgary, Alberta, Canada",155 Mbps
#http://lfs.learnbyexample.com,"Buffalo, New York, USA",155 Mbps

There is a Perl module to parse this kind of database, called
"Text::CSV". It has a built-in validation system that we can use to
report any typo before the site gets updated. I suggest we use the "#"
character to comment out entries. It's more "standard", IMO.

> Anyways, think on it. It will become something to implement in a little
> while.

Some thoughts I had :). Suggestions are welcome.

Anderson Lizardo

More information about the website mailing list